Correlations

• A correlation is the relationship or association between two co-variables.
• A positive correlation is when two co-variables increase together, whereas a negative correlation is when one co-variable increases, the other decreases.
• If there is no relationship between the co-variables it is called a zero correlation.
• What are correlations drawn as?

  scattergrams
• Researchers use statistical tests such as the chi squared test, the Spearman’s rank test or the Pearson’s test to measure the extent of a correlation.
• The extent of a correlational relationship between co-variables is called the correlation coefficient. This has a maximum value of +1 and a minimum of -1.
• +1 is a perfect positive correlation. The closer the correlation coefficient is to this value, the more significant the relationship. Hence the Two co-variables are closely related.
• Strengths of correlations

  + used to analyse trends in data. If significant correlation is found, further investigation is permitted.

  + correlational procedures are easily repeatable.
• Limitations of correlations

  -variables are only measured, not changed so no causal relationships.

  -misinterpretation of correlations as people assume causal relationships.

  -intervening variables may explain the link between the co-variables (shark bites increase as ice cream sales increase, but a sunny day increases swimming and ice cream.)

  -correlations lack internal/external validity & generalisability.